Behavior and Philosophy, 30, 1-13 (2002). © 2002 Cambridge Center for Behavioral Studies RADICAL BEHAVIORISM AND EXCEPTIONAL MEMORY PHENOMENA Rajan Mahadevan John C. Malone University of Tennessee, Knoxville Jon Bailey Florida State University, Tallahassee ABSTRACT: The central claim of this paper is that radical behaviorism and cognitive psychology can both make important contributions to an experimental analysis of a cognitive skill such as memory performance. Though they currently differ in what constitutes an explanation of many phenomena, behaviorists and cognitive psychologists share interests in such human activities as problem solving and memory. We show how the behavioral approach may apply to one case that seems to epitomize cognition—the dramatic improvement in the memory span performance of one individual on a task often used by cognitive psychologists to assess short-term memory. After 230 hours of practice, ability to recall random digits improved from a span of 7 digits to a span of 80. Although a detailed account of the mechanisms that mediated such improvement has been given, we show that the acquisition of such exceptional memory skill can also be explained within the framework of behavior analysis. Key words: exceptional memory, Radical Behaviorism, cognitive explanations, digit span. A Behavioral Analysis of Exceptional Memory Memory has generally been studied by cognitive psychologists and by scientists from some other disciplines, such as the neurosciences. The cognitive movement came about (or at least gained momentum) in the 1950s and 1960s because developments in the field of artificial intelligence, human factors, and psycholinguistics led to the widespread belief that radical behaviorism could not provide a satisfactory account of human behavior. Thus, the spatial metaphors were invented to account for human problem solving, and the structure models were invented to account for human memory phenomena. In contrast to radical behaviorists, cognitive psychologists are more concerned with ordinary-language categories, such as “memory,” “reading comprehension,” “problem solving,” and with determining the limits of human capabilities (e.g., what is the span of immediate memory?). Behavior analysts, on the other hand, are less interested in determining “limits” and more concerned with demonstrating functional relationships. Behaviorists also argue against the use of mediators, such as storage metaphors AUTHORS’ NOTE: Reprints may be obtained from any of the authors. Email: [email protected], [email protected], [email protected]. 1 MAHADEVAN, MALONE, & BAILEY and retrieval mechanisms, mediators that are the essence of cognitive psychology. The widespread view among cognitive psychologists is that behavior analysis does not have any role, or not a significant one, to play in research on human memory. The belief that behavior analysis is excluded (or has excluded itself) from such areas was a major factor in precipitating the so-called cognitive revolution. Those seriously interested in human cognition believed that mediational constructs were indispensable to an understanding of “higher mental functions,” such as human problem solving and memory. However, some, such as Watkins (1990), argue that mediational constructs have impeded progress in memory research. He went so far as to write: Memory theorizing is going nowhere. The reason is that it is rooted in mediationism, the doctrine that memory is mediated by some sort of memory trace. (p. 328) Watkins specifically disputed the usefulness of inferred stages of “encoding,” “storage,” and “retrieval,” arguing that such a model far surpasses the conclusions that may be drawn from the data of experimental psychology. Excessive modeling has led to an overabundance of memory theories that employ terminology that is imprecise and assumptions that are difficult to disprove. Roediger (1985) agreed, claiming that there has been such a proliferation of analogical memory models that a huge array of hypothetical mental representations are tied loosely, if at all, to behavior. The proliferation of hypothetical mechanisms and mediators of memory had been noted earlier by many others, including Underwood (1972) and Tulving (1979). We argue that not only can such “mental” phenomena as memory be the subject of behavior analysis, but also that the behavioral viewpoint may be profitably adopted by cognitive psychologists. We first briefly describe the cognitive and the behavioral approaches to memory. Then we describe a case study of an individual designated as “SF,” who acquired exceptional memory through laboratory practice (Ericsson & Chase, 1982). Third, we provide a cognitive interpretation as well as a behavioral analysis of SF’s performance, and we discuss ways in which behavior analysis can improve our understanding of memory. Finally, we compare and contrast the relative fruitfulness of the radical behaviorist and the cognitive positions, in explaining memory behavior. The Cognitive Approach to Memory Although there are many models of memory, most researchers argue in favor of two distinct memory systems, a short-term memory (STM) and a long-term memory system (LTM). The capacity of STM has been considered to be limited to about seven unrelated items (Miller, 1956) or “chunks,” and to persist over about 30 seconds. After 30 seconds or so information in STM is either “lost” or forgotten unless it has been transferred to more durable storage in LTM. A person can use a number of devices to achieve storage in LTM, such as rehearsal or encoding the material in some meaningful manner, such as chunking. The capacity of LTM is 2 EXCEPTIONAL MEMORY AND RADICAL BEHAVIORISM assumed to be very large and information in LTM can last a lifetime. STM corresponds roughly with the current state of consciousness. Information that is stored in LTM cannot be retrieved directly from it—it must first be brought into STM before the “output” can be produced. There are, of course, many, many models for memory1 and this sketch is an approximation only. Consider it a generic version. Anyone attempting even a brief summary of cognitive approaches to memory must pause and consider the opinion of Endel Tulving (1979, p. 27): After a hundred years of laboratory-based study of memory, we still do not seem to possess any concepts that the majority of workers would consider important or necessary. If one asked a dozen or so randomly selected active memory researchers to compile a list of concepts without which they could not function, one would find little agreement among them . if one compares different textbooks of memory, one discovers that there is little overlap among their subject indexes. It seems that important concepts of one author can apparently be dispensed with by another. Radical behaviorists might argue that matters have not improved since 1979 and many cognitivists might agree! The Behavioral Position on Memory For behaviorists, or radical behaviorists at least, the construct of memory is not required in order to understanding remembering. This position was long expressed by Skinner, for example in Contingencies of Reinforcement (1969), About Behaviorism (1974), and elsewhere. Specifically, Skinner (1969, p. 274) argued that the storage metaphor that cognitive psychology borrowed from computer science is inappropriate, since: . the organism does not need to make a copy of the environment in order to perceive it. When an organism that has been exposed to a set of contingencies of reinforcement is modified by them and so behaves differently, what is stored is a modified organism, not a record of the modifying variables . the conditions that determine the accessibility of stored memories really determine the accessibility of responses. Cognitive psychologists disagree, but the fact is that the metaphor of stored memories is unnecessary to explain behavior. This is not a new nor even a recent view. James Angell (1907) promoted it a century ago, as did many others during the late 19th and early 20th centuries. Baum (1994) follows Skinner as a modern opponent of memory-as-storage and notes that we never say that we “remember” that we stubbed our toe 30 seconds ago. If we don’t need to store-and-retrieve over a 30-second span, we need not do so over a 30-hour or 30-year span. Skinner 1 The controversy about whether one needs to conceptualize memory in terms of one memory system or two (STM and LTM) is irrelevant to the present paper. 3 MAHADEVAN, MALONE, & BAILEY (1974, 1985) proposed that a storage battery was a better model of a behaving organism than is storage in a computer (or other) memory. We store electricity in a battery by changing its chemical composition and we remove electricity when we need it. But there is no electricity in the battery. By the same token, organisms do not acquire various kinds of behavior as possessions; they simply come to act differently in different situations because of past experience. Many current cognitive psychologists would agree with Skinner— the storage metaphor has not proven useful in understanding memory and information is not passively stored (e.g., Ericsson, 1985; Ericsson, Weaver, Delaney, & Mahadevan, 1996). But many believe that the construct of memory is necessary, despite behaviorist objections. One reason that cognitive psychologists see memory as necessary is to preserve the notion of a problem space to deal with cognitive activity such as problem solving, and the spatial metaphor has occupied a pivotal role in many cognitive theories. However, as Rubin (1988) has pointed out, the spatial metaphor predates the cognitive revolution because “Our spatial abilities are so strong and so pervasive that they do not allow for the development of complex process models” (p. 380). This “problem space” provides the static snapshot that seems so necessary to lay readers and to many psychologists. But it lacks a satisfactory description of what actually goes on when the movie begins. A Case Study of Exceptional Memory We turn now to the case study of subject SF. We describe his memory skill acquisition from the cognitive as well as from the behavioral perspectives.